The re-emergence of surgical therapy for Parkinson's disease (PD) has provided major relief for patients who are no longer responsive to medical therapy. Deep brain stimulation (DBS) of either the internal segment of the globus pallidus (GPi) or the subthalamic nucleus (STN) is now being tested as an alternative surgical method to ablation. With DBS, an electrode is placed in the target nucleus and electrical current is applied with a subcutaneous stimulator device. The mechanisms by which DBS alters neural circuitry, culminating in symptom reduction, remains largely unknown. The overall goal of this project is to determine if there are relationships between DBS, the cardinal symptoms of PD, and patterns of brain activity measured at rest and during different types of movement. To do this, brain activity measured with positron tomography (PET) imaging of cerebral blood flow (CBF) will be assessed during GPi and STN DBS in PD patients. The specific aims will examine: 1. the effects of unilateral GPi and STN stimulation on "resting" brain activity. 2. the effects of unilateral GPi and STN stimulation on brain activity during different types of movement. Specific aims 1 and 2 will establish if GPi or STN DBS have similar effects on motor cortical circuits. The tasks are the same as those being used in PD patients studied before and after pallidotomy so that we will also be able to determine if GPi DBS and pallidotomy lead to similar changes of motor circuitry. 3. the clinical and functional (PET) effects of stimulation at different locations within the GPi. 4. the clinical and functional (PET) effect of GPi stimulation as a function of time. Specific aims 3 and 4 are designed to dissociate clinical symptoms with DBS to determine if there are anatomic correlates of the cardinal features of PD. These imaging studies using DBS provide a large-scale analysis of altered neural circuitry in the setting of neurodegeneration and reorganization that cannot be examined at the single neuron level or by behavioral assessment alone. The results of these experiments could have an impact on the development of surgical therapy for Parkinson's disease and could ultimately modify patient or stimulator selection.